[KongchangAI]
Unverified50% confidenceBenchmarkExact time

Le Chonk scored a total of 42/80, or 52.5%, equivalent to 5.25/10, across the eight questions in KingBench3.

1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/11/2026
First Seen
Valid until: 1/9/2027

Sources

Mistral Large 4 Hands-On: How Good Is the "Best Open-Source Model" at Real Programming Tasks?

Related Entities

Le ChonkmodelKingBench3concept

Related Claims

UnverifiedIn the UP host's KingBench 3 eight-item test, GPT-6 Sol scored 66/80 (82.5%) and Opus 5.5 scored 75/80 (93.75%), with Opus leading by 11.25 percentage points.68% similar

Cite This Claim

Stable URI
https://kongchang.com/claim/1000070
API
curl https://kongchang.com/api/v1/knowledge/claims/1000070
MCP
get_claim(id=1000070)